linux-kernel - Re: [PATCH] nvme: determine the number of IO queues

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 18 Apr 2019 14:24:58 +0800
From:   Aaron Ma <aaron.ma@...onical.com>
To:     "Edmund Nadolski (Microsoft)" <edmund.f.nadolski@...il.com>,
        linux-kernel@...r.kernel.org, linux-nvme@...ts.infradead.org,
        keith.busch@...el.com, axboe@...com
Subject: Re: [PATCH] nvme: determine the number of IO queues

On 4/18/19 5:30 AM, Edmund Nadolski (Microsoft) wrote:
> On 4/17/19 7:12 AM, Aaron Ma wrote:
>> Some controllers support limited IO queues, when over set
>> the number, it will return invalid field error.
>> Then NVME will be removed by driver.
>>
>> Find the max number of IO queues that controller supports.
>> When it still got invalid result, set 1 IO queue at least to
>> bring NVME online.
>>
>> Signed-off-by: Aaron Ma <aaron.ma@...onical.com>
>> ---
>>   drivers/nvme/host/core.c | 29 ++++++++++++++++++-----------
>>   1 file changed, 18 insertions(+), 11 deletions(-)
>>
>> diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
>> index 2c43e12b70af..fb7f05c310c8 100644
>> --- a/drivers/nvme/host/core.c
>> +++ b/drivers/nvme/host/core.c
>> @@ -1134,14 +1134,24 @@ static int nvme_set_features(struct nvme_ctrl
>> *dev, unsigned fid, unsigned dword
>>     int nvme_set_queue_count(struct nvme_ctrl *ctrl, int *count)
>>   {
>> -    u32 q_count = (*count - 1) | ((*count - 1) << 16);
>> +    u32 q_count;
>>       u32 result;
>> -    int status, nr_io_queues;
>> -
>> -    status = nvme_set_features(ctrl, NVME_FEAT_NUM_QUEUES, q_count,
>> NULL, 0,
>> -            &result);
>> -    if (status < 0)
>> -        return status;
>> +    int status = -1;
>> +    int nr_io_queues;
>> +    int try_count;
>> +
>> +    for (try_count = *count; try_count > 0; try_count--) {
>> +        q_count = (try_count - 1) | ((try_count - 1) << 16);
> 
> A macro here might help readability.

Will add in V2.

> 
>> +        status = nvme_set_features(ctrl, NVME_FEAT_NUM_QUEUES,
>> +                q_count, NULL, 0, &result);
>> +        if (status < 0)
>> +            return status;
>> +        else if (status == 0) {
> 
> else following return is not needed.

 *      0: successful read
 *      > 0: NVMe error status code
 *      < 0: Linux errno error code

3 conditions should be taken care.
status > 0 will be handled after loop.

else is needed.

> 
> 
>> +            nr_io_queues = min(result & 0xffff, result >> 16) + 1;
> 
> Likewise, a macro as above.

Will add in V2.

> 
> 
> Ed
> 
>> +            *count = min(try_count, nr_io_queues);
>> +            break;
>> +        }
>> +    }
>>         /*
>>        * Degraded controllers might return an error when setting the
>> queue
>> @@ -1150,10 +1160,7 @@ int nvme_set_queue_count(struct nvme_ctrl
>> *ctrl, int *count)
>>        */
>>       if (status > 0) {
>>           dev_err(ctrl->device, "Could not set queue count (%d)\n",
>> status);
>> -        *count = 0;
>> -    } else {
>> -        nr_io_queues = min(result & 0xffff, result >> 16) + 1;
>> -        *count = min(*count, nr_io_queues);
>> +        *count = 1;
>>       }
>>         return 0;
>>
>